A New Matrix-Free Approach for Large-Scale Geodynamic Simulations and its Performance

Beitrag bei einer Tagung


Details zur Publikation

Autor(en): Bauer S, Huber M, Mohr M, Rüde U, Wohlmuth BI
Jahr der Veröffentlichung: 2018
Tagungsband: International Conference on Computational Science
Seitenbereich: 17--30
ISBN: 978-3-319-93700-7
Sprache: Englisch


Abstract

We report on a two-scale approach for efficient matrix-free finite element simulations. The proposed method is based on surrogate element matrices constructed by low-order polynomial approximations. It is applied to a Stokes-type PDE system with variable viscosity as is a key component in mantle convection models. We set the ground for a rigorous performance analysis inspired by the concept of parallel textbook multigrid efficiency and study the weak scaling behavior on SuperMUC, a peta-scale supercomputer system. For a complex geodynamical model, we achieve a parallel efficiency of 95% on up to 47 250 compute cores. Our largest simulation uses a trillion () degrees of freedom for a global mesh resolution of 1.7 km.



FAU-Autoren / FAU-Herausgeber

Bauer, Simon
Lehrstuhl für Informatik 10 (Systemsimulation)
Huber, Markus
Lehrstuhl für Informatik 10 (Systemsimulation)
Rüde, Ulrich Prof. Dr.
Lehrstuhl für Informatik 10 (Systemsimulation)


Autor(en) der externen Einrichtung(en)
Technische Universität München (TUM)


Zitierweisen

APA:
Bauer, S., Huber, M., Mohr, M., Rüde, U., & Wohlmuth, B.I. (2018). A New Matrix-Free Approach for Large-Scale Geodynamic Simulations and its Performance. In International Conference on Computational Science (pp. 17--30).

MLA:
Bauer, Simon, et al. "A New Matrix-Free Approach for Large-Scale Geodynamic Simulations and its Performance." Proceedings of the International Conference on Computational Science 2018. 17--30.

BibTeX: 

Zuletzt aktualisiert 2019-21-03 um 19:10